By Marty Cagan · svpg.com · LinkedIn
Cagan argues that product teams need two distinct disciplines: Product Discovery, which uses prototyping and user testing to find new valuable, usable, feasible solutions, and Product Optimization, which improves an existing product through analytics and A/B testing. Optimization runs as a continuous loop of controlled experiments judged on live data, while Discovery depends on watching customers and making bold pivots. Mixing the two up wastes effort, so the piece matters for product leaders deciding how to spend their time and team capacity.
01Key takeaways
- Use Product Discovery to find new valuable, usable, feasible functionality, and Product Optimization to improve what already exists.
- Instrument your product with analytics before trying to optimize, otherwise improvement ideas are just untested theories.
- Run optimization as a continuous loop of controlled experiments, reviewing results quickly and keeping the best-performing variants.
- Watching users reveals insight but cannot detect small conversion differences; large live-data volumes are needed for that.
- Optimization pairs well with Agile teams because small, frequent changes fit short iteration cycles.
- Don't apply the full Scrum team or optimization-style small changes to discovery work; prototypes and fast pivots suit it better.
02Key sections
- Product Discovery
- Discovery uses high-fidelity prototypes and user testing to find a minimum viable product for significant new functionality or major redesigns.
- Product Optimization
- Optimization improves an existing product's experience and business results using web analytics and A/B testing in a continuous cycle of experiments.
- Instrumentation first
- Without instrumented analytics a team is effectively flying blind, so getting data collection in place is the top priority before optimizing.
- Why the techniques differ
- Discovery favors rapid, daily iterations, bold pivots, and watching customers, while optimization favors small changes measured against large data volumes with a Scrum team.
- Non-web products and skills
- Installed software can gather aggregated usage data and report it back, and strong product leaders should be capable in both disciplines.
03From the post
“A partnership dedicated to teaching best practices to product teams and product leaders”
04Frameworks mentioned
Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.